|
PureBytes Links
Trading Reference Links
|
The Daily-Weekly Stochastic divergence gives interesting
Entries/Exits for futures.
Should we use it as a basis of a trading system to
trade both Light Crude Oil and Gold ?
This method has some suggestions on the subject:
// CLZ03, GCZ03 research, by D. Tsokakis,
Nov2003
START=DateNum()==990802
;Plot(10000,"",1,1);per=100;EVENT=BarsSince(START)%per==0;//
The Daily-Weekly Stochastic DivergenceTimeFrameSet(
inDaily ); sD = StochD(); TimeFrameSet( inWeekly ); sW=StochD();
TimeFrameRestore(); DailyStochD=TimeFrameExpand( sD,
inDaily);WeeklyStochD=TimeFrameExpand(sW,inWeekly);Cond1=Ref(DailyStochD,-1)==HHV(DailyStochD,3);sD1=ValueWhen(Cond1,Ref(DailyStochD,-1));Cond2=Ref(DailyStochD,-1)>70
AND Ref(sD1,-1)>sD1
;Cond3=WeeklyStochD>Ref(WeeklyStochD,-5);Div=Cond2 AND
Cond3;Cond11=Ref(DailyStochD,-1)==LLV(DailyStochD,3);sD11=ValueWhen(Cond11,Ref(DailyStochD,-1));Cond21=Ref(DailyStochD,-1)<30
AND Ref(sD11,-1)<sD11
;Cond31=WeeklyStochD<Ref(WeeklyStochD,-5);Div1=Cond21 AND
Cond31;// The trading system
Buy=div1;Sell=div;Short=Sell;Cover=Buy;// the
Indicator builder
partE34=Equity(1,0);z=Cum(event);Counter=Cum(EVENT)%10;shape=IIf(Counter==1,shapeDigit0,IIf(Counter==2
,shapeDigit1,IIf(Counter==3 ,shapeDigit2,IIf(Counter==4
,shapeDigit3,IIf(Counter==5 ,shapeDigit4,IIf(Counter==6
,shapeDigit5,IIf(Counter==7 ,shapeDigit6,IIf(Counter==8
,shapeDigit7,IIf(Counter==9
,shapeDigit8,shapeDigit9)))))))));PlotShapes(shape*EVENT,colorBlack,0,Graph0,0);dcounter=1+Cum(Counter==1
AND Ref(Counter,-1)==0);dshape=IIf(dCounter==1,shapeNone,IIf(dCounter==2
,shapeDigit1,IIf(dCounter==3 ,shapeDigit2,IIf(dCounter==4
,shapeDigit3,IIf(dCounter==5 ,shapeDigit4,IIf(dCounter==6
,shapeDigit5,IIf(dCounter==7 ,shapeDigit6,IIf(dCounter==8
,shapeDigit7,IIf(dCounter==9
,shapeDigit8,shapeDigit9)))))))));PlotShapes(dshape*event,colorBlack,0,Graph0,10);//
The Automatic Analysis
partAddColumn(z-1,"IP",1.0,colorBlack,colorOrange);AddColumn(E34,"E");for(IPsAgo=2;IPsAgo<LastValue(z);IPsAgo=IPsAgo+1){Progress=100*(-1+E34/ValueWhen(event,E34,IPsAgo+1));AddColumn(Progress,WriteIf(IPsAgo==2,"Progress
","")+WriteVal(IPsAgo,1.0)+" IPs
ago",1.2,colorBlack,IIf(Progress>0,colorBrightGreen,colorRed));}Filter=event;
Explore both above futures for all quotations [data at least
from Jan1999 till now...]
The attached gif is the main purpose of this
study.
The way LightCrudeOil is traded does not match to
this style, it has a lot of problems [red cells] for both the
short-term
and the long-term trader. The Daily-Weekly divergence may give
interesting signals, but it can not be the basis of a system,
its role should be only consultive.
The Gold trading profile is closer to this method.[the matrix
is full of green cells for short and long time frames]
The D-W StochD Div may be the basis of a trading system,
if we suppose that gold traders
will not change dramatically thier habits.
Above all, the Daily-Weekly divergence is a way to "code" the
style of trading.
[CL goes much better with the usual daily chart
Stochastic Divergence, as posted from time to time to this great
list...]
----- Original Message -----
From: <A
title=tsokakis@xxxxxxxxx href="">Dimitris Tsokakis
To: <A title=amibroker@xxxxxxxxxxxxxxx
href="">amibroker@xxxxxxxxxxxxxxx
Sent: Tuesday, November 04, 2003 2:08 PM
Subject: A new method, III
Here is another, interesting, QQQ approach.
The P=15, d=65 gave nice profits the bearish years, has some
small problem the last months of the bearish cycle [IP5 to IP6] and
is profitable again.
The "5 IPs ago" followers [time frame =500 bars] did not even
feel the IP5, IP6 decline.
The respective code is
START=DateNum()==1000601
;Plot(5000,"",1,1);per=100;EVENT=BarsSince(START)%per==0;// The
trading
systemP=Optimize("p",15,5,50,5);X5=100*((Sum(IIf(C>Ref(C,-1),(C-Ref(C,-1)),0),P))-(Sum(IIf(C<Ref(C,-1),(Ref(C,-1)-C),0),P)))/((Sum(IIf(C>Ref(C,-1),(C-Ref(C,-1)),0),P)+(Sum(IIf(C<Ref(C,-1),(Ref(C,-1)-C),0),P))));d=Optimize("d",65,-80,80,1);Sell=Cross(x5,d);Buy=Cross(-D,X5);Buy=ExRem(Buy,Sell);Sell=ExRem(Sell,Buy);Short=Cross(D,X5);Cover=Buy;//
the Indicator builder
partE34=Equity(1,0);z=Cum(event);Counter=Cum(EVENT)%10;shape=IIf(Counter==1,shapeDigit0,IIf(Counter==2
,shapeDigit1,IIf(Counter==3 ,shapeDigit2,IIf(Counter==4
,shapeDigit3,IIf(Counter==5 ,shapeDigit4,IIf(Counter==6
,shapeDigit5,IIf(Counter==7 ,shapeDigit6,IIf(Counter==8
,shapeDigit7,IIf(Counter==9
,shapeDigit8,shapeDigit9)))))))));PlotShapes(shape*EVENT,colorBlack,0,Graph0,0);dcounter=1+Cum(Counter==1
AND Ref(Counter,-1)==0);dshape=IIf(dCounter==1,shapeNone,IIf(dCounter==2
,shapeDigit1,IIf(dCounter==3 ,shapeDigit2,IIf(dCounter==4
,shapeDigit3,IIf(dCounter==5 ,shapeDigit4,IIf(dCounter==6
,shapeDigit5,IIf(dCounter==7 ,shapeDigit6,IIf(dCounter==8
,shapeDigit7,IIf(dCounter==9
,shapeDigit8,shapeDigit9)))))))));PlotShapes(dshape*event,colorBlack,0,Graph0,10);//
The Automatic Analysis
partAddColumn(z-1,"IP",1.0,colorBlack,colorOrange);AddColumn(E34,"E");for(IPsAgo=2;IPsAgo<LastValue(z);IPsAgo=IPsAgo+1){//AddColumn(ValueWhen(event,E34,IPsAgo+1),"E34
2
IPsAgo");Progress=100*(-1+E34/ValueWhen(event,E34,IPsAgo+1));AddColumn(Progress,WriteIf(IPsAgo==2,"Progress
","")+WriteVal(IPsAgo,1.0)+" IPs
ago",1.2,colorBlack,IIf(Progress>0,colorBrightGreen,colorRed));}Filter=event;
Do not be surprised with the modified
d=Optimize("d",65,-80,80,1); , some stocks style is contrarian and it is better
to buy high CMO and sell low...
----- Original Message -----
From: <A
title=tsokakis@xxxxxxxxx href="">Dimitris Tsokakis
To: <A title=amibroker@xxxxxxxxxxxxxxx
href="">amibroker@xxxxxxxxxxxxxxx
Sent: Tuesday, November 04, 2003 1:15 PM
Subject: A new method, II
The same code may give another interesting
result.
Select a stock, say BEAS and hit optimize.
Replace the optimal solution [=41] in the default
place
<FONT
face="Times New Roman">d=<FONT
size=2>Optimize("d"<FONT
size=2>,41,<FONT
size=2>10,80<FONT
size=2>,1);
and Explore the current stock for all quotations.
The result matrix is quite informative for the past behavior
of the recent optimal solution. It was always good, had some problems around
IP5, IP6
[autumn 2002] but came back to profits.
The optimal d=64 for CSCO was not
always safe.
For INTC, the d=13 is also interesting : After a congestive
period around IP5, IP6 [spring-autumn 2002] it came back to progressive
profits.
The optimal d=54 for ^NDX is rather coincidental [the profits
are mainly due to the period IP2 to IP3] and does not give [IMO] any kind of
guarantee for a future use.
----- Original Message -----
From: <A
title=tsokakis@xxxxxxxxx href="">Dimitris Tsokakis
To: <A title=amibroker@xxxxxxxxxxxxxxx
href="">amibroker@xxxxxxxxxxxxxxx
Sent: Tuesday, November 04, 2003 12:28 PM
Subject: A new method
Except the Equity values, we need the Equity progress as time
goes by.
A stock may be highly profitable for 6 months, two years
ago and presents high Equity value in a 3-year period,
but it may be a loser the last year.
For this study we may establish some Inspection Points,
equally distributed in the test period and check the relative Equity
progress.
The code is divided into 4 parts.
1. Introduction
START=DateNum()==1000601 ;Plot(5000,"",1,1);
My data begin on Jan2000. The majority of the
indicators I use are well defined the first 6 months. June1, 2000 is a good
starting point .
The 5000 line is placed below the initial Equity 10000, many
stocks have loosing periods and it is used to align the IP digits.
per=100;EVENT=BarsSince(START)%per==0;The IPs will be
placed every 100 bars for the inspection EVENT. The distance of each IP from the
starting date is a multiple of 100.
2. The trading system
CMO5 with threshold d=58 was selected for the
examples.Long&Short trades are included, the system is almost always in the
market.
3. The Indicator builder
part
Scan and backtest the stock [or the database] for all
quotations. Hit Equity button and you will automatically have a new IB pane
with
the Equity, the B&H yellow line and the IPs counter on the
same graph. If you place your cursor on IP6 you may read the respective
Equity
and B&H value.
4. The AA part.
Explore the stock [or the database] for all
quotations.
The
Filter=event;
will check only the IP dates.
The 1st column is the name of the stock, the 2nd is the IP
dates
The 3rd column is the IP number [from 0 to 8 in the
example]
The 4th column[E34] is the Equity value for the respective
IP.
We see for APOL 21,830 on 1/6/2000, 28900 on 23/10/2000
etc.
The 5th column[Progress 2 IPs ago] checks per IP the % of the
Equity progress 2 IPs ago.
The 8th IP is compared to the 6th, the 7th is compared to the
5th etc.
The two first IPs are without comparison.
The 6th column[3 IPS ago] checks per IP the % progress of
the Equity compared to the IP 3 IPs ago.
The comparison will be continued
until the progress of the last IP related to the first.
We read in APOL last row, last column 314.64. The Equity
for IP8 is 90,518.73, the Equity for IP0 is 21,830.66
and the progress is 90,518.73/21,830.66=4.1464 or 314.64%,
which gives the gain if the Entry is IP0 and the Exit is IP8.
For easy reading, positive progress has green background and
negative progress has red.
// The progress of the Equity curve, by D.
Tsokakis, Nov2003
START=DateNum()==1000601
;Plot(5000,"",1,1);per=100;EVENT=BarsSince(START)%per==0;//
The trading
systemP=5;X5=100*((Sum(IIf(C>Ref(C,-1),(C-Ref(C,-1)),0),P))-(Sum(IIf(C<Ref(C,-1),(Ref(C,-1)-C),0),P)))/((Sum(IIf(C>Ref(C,-1),(C-Ref(C,-1)),0),P)+(Sum(IIf(C<Ref(C,-1),(Ref(C,-1)-C),0),P))));d=Optimize("d",58,10,80,1);Sell=Cross(x5,d);Buy=Cross(-D,X5);Buy=ExRem(Buy,Sell);Sell=ExRem(Sell,Buy);Short=Cross(D,X5);Cover=Buy;//
the Indicator builder
partE34=Equity(1,0);z=Cum(event);Counter=Cum(EVENT)%10;shape=IIf(Counter==1,shapeDigit0,IIf(Counter==2
,shapeDigit1,IIf(Counter==3 ,shapeDigit2,IIf(Counter==4
,shapeDigit3,IIf(Counter==5 ,shapeDigit4,IIf(Counter==6
,shapeDigit5,IIf(Counter==7 ,shapeDigit6,IIf(Counter==8
,shapeDigit7,IIf(Counter==9
,shapeDigit8,shapeDigit9)))))))));PlotShapes(shape*EVENT,colorBlack,0,Graph0,0);dcounter=1+Cum(Counter==1
AND Ref(Counter,-1)==0);dshape=IIf(dCounter==1,shapeNone,IIf(dCounter==2
,shapeDigit1,IIf(dCounter==3 ,shapeDigit2,IIf(dCounter==4
,shapeDigit3,IIf(dCounter==5 ,shapeDigit4,IIf(dCounter==6
,shapeDigit5,IIf(dCounter==7 ,shapeDigit6,IIf(dCounter==8
,shapeDigit7,IIf(dCounter==9
,shapeDigit8,shapeDigit9)))))))));PlotShapes(dshape*event,colorBlack,0,Graph0,10);//
The Automatic Analysis
partAddColumn(z-1,"IP",1.0,colorBlack,colorOrange);
<FONT
size=2>AddColumn(E34,"E34");for(IPsAgo=2;IPsAgo<LastValue(z);IPsAgo=IPsAgo+1){Progress=100*(-1+E34/ValueWhen(event,E34,IPsAgo+1));AddColumn(Progress,WriteIf(IPsAgo==2,"Progress
","")+WriteVal(IPsAgo,1.0)+" IPs
ago",1.2,colorBlack,IIf(Progress>0,colorBrightGreen,colorRed));}Filter=event;
In every cell of this AA matrix we read the % progress of the
Equity in comparison to the previous IPs [in the example per=100 means we check
200 bars ago, 300 bars ago etc]
Let us have a first taste of the matrix info:
*APOL is full of green cells. At any
IP the correlation to any previous IP is positive. The first column presents the
progress related to 2IPs ago.
It was +129.10% back on March2001, +58.16% <FONT
size=2>on Oct2002 and just a +5.19% on Aug2003, compared to the Equity 200 bars
ago.
[Time for a red cell in the next inspection ???]
*The tough question for APOL is already a fact for
BEAS.
The last IP has red cells, compared to the 2IPs ago, 3IPS ago,
4IPS ago and 5IPS ago [500 bars ago]
*APCC was a horrible example for CMO5,
it was steadily a loser from IP0 to IP6.
The last IP8 is 97.12% profitable for the last 200 bars and
34.86% profitable for the last 300 bars.
The preference of APOL or APCC is not an easy question.
A recent trader would enjoy almost a double price with APCC
and only a 5% with the top APOL.
*BGEN has some CMO5 problems the last
3 IPs. The small +3.24% [IP5, column 2IPs ago] was quite informative for BGEN
future movements...
*INTC is not good the last 400 bars
[The small +3.83 at the IP4 anticipated a red future...]
The last -9.50% is an indication for positive IPs
soon.
*INTU has some good CMO5 news. After
the decline of the "2IPs ago" column [66%, 39%, 13%, 2%] a new +46% came to
change the perspectives .
I think we may "read" a lot of informations from this
matrix.
Dimitris Tsokakis
Yahoo! Groups Sponsor
ADVERTISEMENT
Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
Attachment:
1futures.gif
Attachment:
Description: "Description: GIF image"
|